Graceful Shutdown in Kubernetes: Handling SIGTERM Without Dropped Requests
Learn why pods drop requests on SIGTERM, and how a preStop delay, readiness flip and bounded drain inside the grace period give zero downtime deploys.
14 min read
Tag
Learn why pods drop requests on SIGTERM, and how a preStop delay, readiness flip and bounded drain inside the grace period give zero downtime deploys.
14 min read
Learn how Kubernetes graceful shutdown really works, why endpoint removal races SIGTERM, and how to size preStop hooks and grace periods for zero-error deploys.
14 min read
Why pod shutdown causes 502s, and how preStop delays, terminationGracePeriodSeconds budgets and app-side draining fix the endpoint propagation race.
14 min read
Why server.shutdown=graceful alone drops requests on Kubernetes, and how a preStop delay and a timeout budget give you zero-downtime pod termination.
13 min read